Launched this week
AB-Labz
Plan, analyze, and manage A/B tests in one workspace
7 followers
Plan, analyze, and manage A/B tests in one workspace
7 followers
AB-Labz is a lightweight experimentation workspace for analysts and product teams. Upload a dataset and get a complete A/B or multi-variant test analysis in minutes — no SDK, warehouse integration, or engineering support required. Add the hypothesis, design, conclusions, and decision, export the result to Markdown, and build a searchable experiment archive over time. Plan tests, forecast outcomes, and track program impact when you need more.
Hi Product Hunt! I’m Petr, a product analyst and the founder of AB-Labz.
AB-Labz started as a tool we built for ourselves. We wanted a faster way to analyze experiments, keep their context, and avoid rebuilding the same workflow across SQL exports, spreadsheets, notebooks, and project trackers. It was never meant to be a heavy corporate platform — just something practical that made our own work easier. We use it every day, and we hope it will be useful to you too.
AB-Labz is a lightweight experimentation workspace that starts with the dataset you already have.
You can register, upload a prepared dataset, and get a complete A/B or multi-variant test analysis in minutes — without installing an SDK, connecting a warehouse, changing your traffic allocation system, or involving an engineering team. Even if your company is not ready to adopt a new tool, you can use AB-Labz as your personal statistical layer alongside Jira, Confluence, notebooks, and whatever else your team already uses.
The statistical engine was developed together with two practicing statistics professors, so it does more than run a basic significance test. It checks the data, selects suitable methods, and automatically handles skewed, noisy, and otherwise messy real-world distributions.
Then you can add the hypothesis, design, conclusions, and final decision, keep everything as a structured experiment record, export it to Markdown, or share the result through a public link that anyone can open without an account.
Over time, those records become a searchable experiment archive. When you need more structure, AB-Labz also includes sample-size planning, forecasting, Bayesian monitoring, hypothesis management, Kanban, Gantt, calendar planning, and portfolio-level analytics that show how the entire experimentation program is performing and what it delivers to the business.
AB-Labz does not replace GrowthBook, Statsig, Optimizely, or another feature-flagging and traffic-splitting tool. It works alongside them — or with your internal assignment system.
There is a free Solo plan for individual analysts.
I’d really appreciate your feedback: what is currently the most painful part of your experimentation workflow?